The previous fix for the circlular lock splat about the busyness
worker wasn't quite complete. Even though the reset-in-progress flag
is cleared at the start of intel_uc_reset_finish, the entire function
is still inside the reset mutex lock. Not sure why the patch appeared
to fix the issue both locally and in CI. However, it is now back
again.
There is a further complication that the wedge code path within
intel_gt_reset() jumps around so much that it results in nested
reset_prepare/_finish calls. That is, the call sequence is:
intel_gt_reset
| reset_prepare
| __intel_gt_set_wedged
| | reset_prepare
| | reset_finish
| reset_finish
The nested finish means that even if the clear of the in-progress flag
was moved to the end of _finish, it would still be clear for the
entire second call. Surprisingly, this does not seem to be causing any
other problems at present.
As an aside, a wedge on fini does not call the finish functions at
all. The reset_in_progress flag is left set (twice).
So instead of trying to cancel the worker anywhere at all in the reset
path, just add a cancel to intel_guc_submission_fini instead. Note
that it is not a problem if the worker is still active during a reset.
Either it will run before the reset path starts locking things and
will simply block the reset code for a tiny amount of time. Or it will
run after the locks have been acquired and will early exit due to the
try-lock.
Also, do not use the reset-in-progress flag to decide whether a
synchronous cancel is safe (from a lockdep perspective) or not.
Instead, use the actual reset mutex state (both the genuine one and
the custom rolled BACKOFF one).
Fixes: 0e00a8814e ("drm/i915/guc: Avoid circular locking issue on busyness flush")
Signed-off-by: John Harrison <John.C.Harrison@Intel.com>
Cc: Zhanjun Dong <zhanjun.dong@intel.com>
Cc: John Harrison <John.C.Harrison@Intel.com>
Cc: Andi Shyti <andi.shyti@linux.intel.com>
Cc: Daniel Vetter <daniel@ffwll.ch>
Cc: Daniel Vetter <daniel.vetter@ffwll.ch>
Cc: Rodrigo Vivi <rodrigo.vivi@intel.com>
Cc: Nirmoy Das <nirmoy.das@intel.com>
Cc: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Cc: Umesh Nerlige Ramappa <umesh.nerlige.ramappa@intel.com>
Cc: Andrzej Hajda <andrzej.hajda@intel.com>
Cc: Matt Roper <matthew.d.roper@intel.com>
Cc: Jonathan Cavitt <jonathan.cavitt@intel.com>
Cc: Prathap Kumar Valsan <prathap.kumar.valsan@intel.com>
Cc: Alan Previn <alan.previn.teres.alexis@intel.com>
Cc: Madhumitha Tolakanahalli Pradeep <madhumitha.tolakanahalli.pradeep@intel.com>
Cc: Daniele Ceraolo Spurio <daniele.ceraolospurio@intel.com>
Cc: Ashutosh Dixit <ashutosh.dixit@intel.com>
Cc: Dnyaneshwar Bhadane <dnyaneshwar.bhadane@intel.com>
Reviewed-by: Nirmoy Das <nirmoy.das@intel.com>
Reviewed-by: Andi Shyti <andi.shyti@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20240329235306.1559639-1-John.C.Harrison@Intel.com
(cherry picked from commit 3563d85531)
Signed-off-by: Rodrigo Vivi <rodrigo.vivi@intel.com>
The context persistence code does things like send super high priority
heartbeat pulses to ensure any leaked context can still be pre-empted
and thus isn't a total denial of service but only a minor denial of
service. Unfortunately, it wasn't bothering to restart the heartbeat
worker with a fresh timeout. Thus, if a persistent context happened to
be closed just before the heartbeat was going to go ping anyway then
the forced pulse would get a negligble execution time. And as the
forced pulse is super high priority, the worker thread's next step is
a reset. Which means a potentially innocent system randomly goes boom
when attempting to close a context. So, force a re-schedule of the
worker thread with the appropriate timeout.
Signed-off-by: John Harrison <John.C.Harrison@Intel.com>
Reviewed-by: Andi Shyti <andi.shyti@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20240110210216.4125092-1-John.C.Harrison@Intel.com
Now that the GGTT PTE updates go straight to GSMBASE (bypassing
GTTMMADR) there should be no more risk of system hangs? So the
"binder" (ie. update the PTEs via MI_UPDATE_GTT) is no longer
necessary, disable it.
My main worry with the MI_UPDATE_GTT are:
- only used on this one platform so very limited testing coverage
- async so more opprtunities to screw things up
- what happens if the engine hangs while we're waiting for MI_UPDATE_GTT
to finish?
- requires working command submission, so even getting a working
display now depends on a lot more extra components working correctly
TODO: MI_UPDATE_GTT might be interesting as an optimization
though, so perhaps someone should look into always using it
(assuming the GPU is alive and well)?
v2: Keep using MI_UPDATE_GTT on VM guests
v3: use i915_direct_stolen_access()
Reviewed-by: Nirmoy Das <nirmoy.das@intel.com>
Tested-by: Paz Zcharya <pazz@chromium.org>
Signed-off-by: Ville Syrjälä <ville.syrjala@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20240202224340.30647-6-ville.syrjala@linux.intel.com
On MTL accessing stolen memory via the BARs is somehow borked,
and it can hang the machine. As a workaround let's bypass the
BARs and just go straight to DSMBASE/GSMBASE instead.
Note that on every other platform this itself would hang the
machine, but on MTL the system firmware is expected to relax
the access permission guarding stolen memory to enable this
workaround, and thus direct CPU accesses should be fine.
The raw stolen memory areas won't be passed to VMs so we'll
need to risk using the BAR there for the initial setup. Once
command submission is up we should switch to MI_UPDATE_GTT
which at least shouldn't hang the whole machine.
v2: Don't use direct GSM/DSM access on guests
Add w/a number
v3: Check register 0x138914 to see if pcode did its job
Add some debug prints
Cc: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
Reviewed-by: Andrzej Hajda <andrzej.hajda@intel.com>
Reviewed-by: Radhakrishna Sripada <radhakrishna.sripada@intel.com>
Reviewed-by: Nirmoy Das <nirmoy.das@intel.com>
Tested-by: Paz Zcharya <pazz@chromium.org>
Signed-off-by: Ville Syrjälä <ville.syrjala@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20240202224340.30647-5-ville.syrjala@linux.intel.com
mem->region is a struct resource, but mem->io_start and
mem->io_size are not for whatever reason. Let's unify this
and convert the io stuff into a struct resource as well.
Should make life a little less annoying when you don't have
juggle between two different approaches all the time.
Mostly done using cocci (with manual tweaks at all the
places where we mutate io_size by hand):
@@
struct intel_memory_region *M;
expression START, SIZE;
@@
- M->io_start = START;
- M->io_size = SIZE;
+ M->io = DEFINE_RES_MEM(START, SIZE);
@@
struct intel_memory_region *M;
@@
- M->io_start
+ M->io.start
@@
struct intel_memory_region M;
@@
- M.io_start
+ M.io.start
@@
expression M;
@@
- M->io_size
+ resource_size(&M->io)
@@
expression M;
@@
- M.io_size
+ resource_size(&M.io)
Reviewed-by: Andrzej Hajda <andrzej.hajda@intel.com>
Acked-by: Nirmoy Das <nirmoy.das@intel.com>
Tested-by: Paz Zcharya <pazz@chromium.org>
Signed-off-by: Ville Syrjälä <ville.syrjala@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20240202224340.30647-2-ville.syrjala@linux.intel.com
Currently if rc6 is supported, it gets enabled and the sysfs files for
rc6_enable_show and rc6_enable_dev_show uses masks to check information
from drm_i915_private.
However rc6_support functions take more variables and conditions into
consideration and thus these masks are not enough for most of the modern
hardware and it is simpley lyting to the user.
Let's fix it by at least use the rc6.supported flag from intel_gt
information.
Signed-off-by: Juan Escamilla <jcescami@wasd.net>
Reviewed-by: Rodrigo Vivi <rodrigo.vivi@intel.com>
Signed-off-by: Rodrigo Vivi <rodrigo.vivi@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20240110010302.553597-1-jcescami@wasd.net
Document nested struct members with full names as described in
Documentation/doc-guide/kernel-doc.rst.
intel_guc.h:305: warning: Excess struct member 'lock' description in 'intel_guc'
intel_guc.h:305: warning: Excess struct member 'guc_ids' description in 'intel_guc'
intel_guc.h:305: warning: Excess struct member 'num_guc_ids' description in 'intel_guc'
intel_guc.h:305: warning: Excess struct member 'guc_ids_bitmap' description in 'intel_guc'
intel_guc.h:305: warning: Excess struct member 'guc_id_list' description in 'intel_guc'
intel_guc.h:305: warning: Excess struct member 'guc_ids_in_use' description in 'intel_guc'
intel_guc.h:305: warning: Excess struct member 'destroyed_contexts' description in 'intel_guc'
intel_guc.h:305: warning: Excess struct member 'destroyed_worker' description in 'intel_guc'
intel_guc.h:305: warning: Excess struct member 'reset_fail_worker' description in 'intel_guc'
intel_guc.h:305: warning: Excess struct member 'reset_fail_mask' description in 'intel_guc'
intel_guc.h:305: warning: Excess struct member 'sched_disable_delay_ms' description in 'intel_guc'
intel_guc.h:305: warning: Excess struct member 'sched_disable_gucid_threshold' description in 'intel_guc'
intel_guc.h:305: warning: Excess struct member 'lock' description in 'intel_guc'
intel_guc.h:305: warning: Excess struct member 'gt_stamp' description in 'intel_guc'
intel_guc.h:305: warning: Excess struct member 'ping_delay' description in 'intel_guc'
intel_guc.h:305: warning: Excess struct member 'work' description in 'intel_guc'
intel_guc.h:305: warning: Excess struct member 'shift' description in 'intel_guc'
intel_guc.h:305: warning: Excess struct member 'last_stat_jiffies' description in 'intel_guc'
18 warnings as Errors
Signed-off-by: Randy Dunlap <rdunlap@infradead.org>
Cc: Jani Nikula <jani.nikula@linux.intel.com>
Cc: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
Cc: Rodrigo Vivi <rodrigo.vivi@intel.com>
Cc: Tvrtko Ursulin <tvrtko.ursulin@linux.intel.com>
Cc: intel-gfx@lists.freedesktop.org
Cc: Jonathan Corbet <corbet@lwn.net>
Cc: dri-devel@lists.freedesktop.org
Reviewed-by: Rodrigo Vivi <rodrigo.vivi@intel.com>
Reviewed-by: Andi Shyti <andi.shyti@linux.intel.com>
Signed-off-by: Andi Shyti <andi.shyti@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20231226195432.10891-3-rdunlap@infradead.org
(cherry picked from commit e4cf1a70fa)
Signed-off-by: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
Avoid the following lockdep complaint:
<4> [298.856498] ======================================================
<4> [298.856500] WARNING: possible circular locking dependency detected
<4> [298.856503] 6.7.0-rc5-CI_DRM_14017-g58ac4ffc75b6+ #1 Tainted: G
N
<4> [298.856505] ------------------------------------------------------
<4> [298.856507] kworker/4:1H/190 is trying to acquire lock:
<4> [298.856509] ffff8881103e9978 (>->reset.backoff_srcu){++++}-{0:0}, at:
_intel_gt_reset_lock+0x35/0x380 [i915]
<4> [298.856661]
but task is already holding lock:
<4> [298.856663] ffffc900013f7e58
((work_completion)(&(&guc->timestamp.work)->work)){+.+.}-{0:0}, at:
process_scheduled_works+0x264/0x530
<4> [298.856671]
which lock already depends on the new lock.
The complaint is not actually valid. The busyness worker thread does
indeed hold the worker lock and then attempt to acquire the reset lock
(which may have happened in reverse order elsewhere). However, it does
so with a trylock that exits if the reset lock is not available
(specifically to prevent this and other similar deadlocks).
Unfortunately, lockdep does not understand the trylock semantics (the
lock is an i915 specific custom implementation for resets).
Not doing a synchronous flush of the worker thread when a reset is in
progress resolves the lockdep splat by never even attempting to grab
the lock in this particular scenario.
There are situatons where a synchronous cancel is required, however.
So, always do the synchronous cancel if not in reset. And add an extra
synchronous cancel to the end of the reset flow to account for when a
reset is occurring at driver shutdown and the cancel is no longer
synchronous but could lead to unallocated memory accesses if the
worker is not stopped.
Signed-off-by: Zhanjun Dong <zhanjun.dong@intel.com>
Signed-off-by: John Harrison <John.C.Harrison@Intel.com>
Cc: Andi Shyti <andi.shyti@linux.intel.com>
Cc: Daniel Vetter <daniel@ffwll.ch>
Reviewed-by: Andi Shyti <andi.shyti@linux.intel.com>
Acked-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Link: https://patchwork.freedesktop.org/patch/msgid/20231219195957.212600-1-John.C.Harrison@Intel.com
If we are at the end of suspend or very early in resume
its possible an async fence signal (via rcu_call) is triggered
to free_engines which could lead us to the execution of
the context destruction worker (after a prior worker flush).
Thus, when suspending, insert rcu_barriers at the start
of i915_gem_suspend (part of driver's suspend prepare) and
again in i915_gem_suspend_late so that all such cases have
completed and context destruction list isn't missing anything.
In destroyed_worker_func, close the race against CT-loss
by checking that CT is enabled before calling into
deregister_destroyed_contexts.
Based on testing, guc_lrc_desc_unpin may still race and fail
as we traverse the GuC's context-destroy list because the
CT could be disabled right before calling GuC's CT send function.
We've witnessed this race condition once every ~6000-8000
suspend-resume cycles while ensuring workloads that render
something onscreen is continuously started just before
we suspend (and the workload is small enough to complete
and trigger the queued engine/context free-up either very
late in suspend or very early in resume).
In such a case, we need to unroll the entire process because
guc-lrc-unpin takes a gt wakeref which only gets released in
the G2H IRQ reply that never comes through in this corner
case. Without the unroll, the taken wakeref is leaked and will
cascade into a kernel hang later at the tail end of suspend in
this function:
intel_wakeref_wait_for_idle(>->wakeref)
(called by) - intel_gt_pm_wait_for_idle
(called by) - wait_for_suspend
Thus, do an unroll in guc_lrc_desc_unpin and deregister_destroyed_-
contexts if guc_lrc_desc_unpin fails due to CT send falure.
When unrolling, keep the context in the GuC's destroy-list so
it can get picked up on the next destroy worker invocation
(if suspend aborted) or get fully purged as part of a GuC
sanitization (end of suspend) or a reset flow.
Signed-off-by: Alan Previn <alan.previn.teres.alexis@intel.com>
Signed-off-by: Anshuman Gupta <anshuman.gupta@intel.com>
Tested-by: Mousumi Jana <mousumi.jana@intel.com>
Acked-by: Daniele Ceraolo Spurio <daniele.ceraolospurio@intel.com>
Reviewed-by: Rodrigo Vivi <rodrigo.vivi@intel.com>
Signed-off-by: Matt Roper <matthew.d.roper@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20231229215143.581619-1-alan.previn.teres.alexis@intel.com
A failure to load the HuC is occasionally observed where the cause is
believed to be a low GT frequency leading to very long load times.
So a) increase the timeout so that the user still gets a working
system even in the case of slow load. And b) report the frequency
during the load to see if that is the cause of the slow down.
Also update the similar code on the GuC load to not use uncore->gt
when there is a local gt available. The two should match, but no need
for unnecessary de-referencing.
Signed-off-by: John Harrison <John.C.Harrison@Intel.com>
Reviewed-by: Daniele Ceraolo Spurio <daniele.ceraolospurio@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20240102222202.310495-1-John.C.Harrison@Intel.com
Document nested struct members with full names as described in
Documentation/doc-guide/kernel-doc.rst.
intel_guc.h:305: warning: Excess struct member 'lock' description in 'intel_guc'
intel_guc.h:305: warning: Excess struct member 'guc_ids' description in 'intel_guc'
intel_guc.h:305: warning: Excess struct member 'num_guc_ids' description in 'intel_guc'
intel_guc.h:305: warning: Excess struct member 'guc_ids_bitmap' description in 'intel_guc'
intel_guc.h:305: warning: Excess struct member 'guc_id_list' description in 'intel_guc'
intel_guc.h:305: warning: Excess struct member 'guc_ids_in_use' description in 'intel_guc'
intel_guc.h:305: warning: Excess struct member 'destroyed_contexts' description in 'intel_guc'
intel_guc.h:305: warning: Excess struct member 'destroyed_worker' description in 'intel_guc'
intel_guc.h:305: warning: Excess struct member 'reset_fail_worker' description in 'intel_guc'
intel_guc.h:305: warning: Excess struct member 'reset_fail_mask' description in 'intel_guc'
intel_guc.h:305: warning: Excess struct member 'sched_disable_delay_ms' description in 'intel_guc'
intel_guc.h:305: warning: Excess struct member 'sched_disable_gucid_threshold' description in 'intel_guc'
intel_guc.h:305: warning: Excess struct member 'lock' description in 'intel_guc'
intel_guc.h:305: warning: Excess struct member 'gt_stamp' description in 'intel_guc'
intel_guc.h:305: warning: Excess struct member 'ping_delay' description in 'intel_guc'
intel_guc.h:305: warning: Excess struct member 'work' description in 'intel_guc'
intel_guc.h:305: warning: Excess struct member 'shift' description in 'intel_guc'
intel_guc.h:305: warning: Excess struct member 'last_stat_jiffies' description in 'intel_guc'
18 warnings as Errors
Signed-off-by: Randy Dunlap <rdunlap@infradead.org>
Cc: Jani Nikula <jani.nikula@linux.intel.com>
Cc: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
Cc: Rodrigo Vivi <rodrigo.vivi@intel.com>
Cc: Tvrtko Ursulin <tvrtko.ursulin@linux.intel.com>
Cc: intel-gfx@lists.freedesktop.org
Cc: Jonathan Corbet <corbet@lwn.net>
Cc: dri-devel@lists.freedesktop.org
Reviewed-by: Rodrigo Vivi <rodrigo.vivi@intel.com>
Reviewed-by: Andi Shyti <andi.shyti@linux.intel.com>
Signed-off-by: Andi Shyti <andi.shyti@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20231226195432.10891-3-rdunlap@infradead.org
The use of kmap_atomic() is being deprecated in favor of
kmap_local_page()[1], and this patch converts the call from
kmap_atomic() to kmap_local_page().
The main difference between atomic and local mappings is that local
mappings doesn't disable page faults or preemption (the preemption is
disabled for !PREEMPT_RT case, otherwise it only disables migration).
With kmap_local_page(), we can avoid the often unwanted side effect of
unnecessary page faults or preemption disables.
In drm/i915/gt/uc/intel_us_fw.c, the function intel_uc_fw_copy_rsa()
just use the mapping to do memory copy so it doesn't need to disable
pagefaults and preemption for mapping. Thus the local mapping without
atomic context (not disable pagefaults / preemption) is enough.
Therefore, intel_uc_fw_copy_rsa() is a function where the use of
memcpy_from_page() with kmap_local_page() in place of memcpy() with
kmap_atomic() is correctly suited.
Convert the calls of memcpy() with kmap_atomic() / kunmap_atomic() to
memcpy_from_page() which uses local mapping to copy.
[1]: https://lore.kernel.org/all/20220813220034.806698-1-ira.weiny@intel.com/T/#u
Suggested-by: Ira Weiny <ira.weiny@intel.com>
Suggested-by: Fabio M. De Francesco <fmdefrancesco@gmail.com>
Signed-off-by: Zhao Liu <zhao1.liu@intel.com>
Reviewed-by: Ira Weiny <ira.weiny@intel.com>
Reviewed-by: Fabio M. De Francesco <fmdefrancesco@gmail.com>
Signed-off-by: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20231203132947.2328805-8-zhao1.liu@linux.intel.com
Commit 503579448d ("drm/i915/gsc: Mark internal GSC engine with reserved uabi class")
made the GSC0 engine not have a valid uabi class and so broke the engine
reset counting, which in turn was made class based in cb823ed991 ("drm/i915/gt: Use intel_gt as the primary object for handling resets").
Despite the title and commit text of the latter is not mentioning it (and
has left the storage array incorrectly sized), tracking by class, despite
it adding aliasing in hypthotetical multi-tile systems, is handy for
virtual engines which for instance do not have a valid engine->id.
Therefore we keep that but just change it to use the internal class which
is always valid. We also add a helper to increment the count, which
aligns with the existing getter.
What was broken without this fix were out of bounds reads every time a
reset would happen on the GSC0 engine, or during selftests when storing
and cross-checking the counts in igt_live_test_begin and
igt_live_test_end.
Signed-off-by: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Fixes: dfed6b58d5 ("drm/i915/gsc: Mark internal GSC engine with reserved uabi class")
[tursulin: fixed Fixes tag]
Reported-by: Alan Previn Teres Alexis <alan.previn.teres.alexis@intel.com>
Cc: Daniele Ceraolo Spurio <daniele.ceraolospurio@intel.com>
Reviewed-by: Daniele Ceraolo Spurio <daniele.ceraolospurio@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20231201122109.729006-2-tvrtko.ursulin@linux.intel.com
Noticed that the hangcheck selftest is submitting a non-preemptoble
spinner. That means that even if the GuC does not die, the heartbeat
will still kick in and trigger a reset. Which is rather defeating the
purpose of the test - to verify that the heartbeat will kick in if the
GuC itself has died. The test is deliberately killing the GuC, so it
should never hit the case of a non-dead GuC. But it is not impossible
that the kill might fail at some future point due to other driver
re-work.
So, make the spinner pre-emptible. That way the heartbeat can get
through if the GuC is alive and context switching. Thus a reset only
happens if the GuC dies. Thus, if the kill should stop working the
test will now fail rather than claim to pass.
Signed-off-by: John Harrison <John.C.Harrison@Intel.com>
Reviewed-by: Daniele Ceraolo Spurio <daniele.ceraolospurio@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20231114010016.234570-2-John.C.Harrison@Intel.com
The GSC CS is not exposed to the user, so we skipped assigning a uabi
class number for it. However, the trace logs use the uabi class and
instance to identify the engine, so leaving uabi class unset makes the
GSC CS show up as the RCS in those logs.
Given that the engine is not exposed to the user, we can't add a new
case in the uabi enum, so we insted internally define a kernel
internal class as -1.
At the same time remove special handling for the name and complete
the uabi_classes array so internal class is automatically correctly
assigned.
Engine will show as 65535:0 other0 in the logs/traces which should
be unique enough.
v2:
* Fix uabi class u8 vs u16 type confusion.
Signed-off-by: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Fixes: 194babe26b ("drm/i915/mtl: don't expose GSC command streamer to the user")
Cc: Daniele Ceraolo Spurio <daniele.ceraolospurio@intel.com>
Cc: Alan Previn <alan.previn.teres.alexis@intel.com>
Cc: Matt Roper <matthew.d.roper@intel.com>
Reviewed-by: Daniele Ceraolo Spurio <daniele.ceraolospurio@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20231116084456.291533-1-tvrtko.ursulin@linux.intel.com
(cherry picked from commit dfed6b58d5)
Signed-off-by: Jani Nikula <jani.nikula@intel.com>
